A comprehensive index of artificial intelligence and machine-learning research with AI-generated summaries, citation metrics, and direct links to papers and code.
Komal Kumar, Tajamul Ashraf, Omkar Thawakar, et al.
A systematic survey of post-training techniques for LLMs, covering fine-tuning, reinforcement learning, and test-time scaling, with a public repository for tracking developments.
Guanghao Ye, Khiem Pham, Xinzhi Zhang, et al.
Proposes RLSP, a post-training framework that decouples exploration and correctness signals during PPO to enable emergent reasoning behaviors in LLMs.
Keming Wu, Baoyi Wang, Kaichen Zhang, et al.
StreamOPD is a post-training recipe combining verifiable streaming-video data, thinking-mode on-policy distillation, and instruct-mode deployment, improving streaming video understanding without inference-time memory or retrieval.
Xiaoyu Zhu, Xinke Deng, Suresh Taddewadikar, et al.
Introduces Internalized Visual Thinking (IVT), a post-training framework that predicts latent future frames during training to enable efficient, direct-answer video reasoning at inference, outperforming Visual CoT with 5x lower latency.
He-Yen Hsieh, H. T. Kung
ReRound is a post-training quantization method that uses a conditional diffusion model to resolve midpoint ambiguity in round-to-nearest, improving low-bit LLM quantization without calibration data.
Unknown
Introduces RIPT-VLA, a reinforcement-learning-based interactive post-training paradigm for fine-tuning pretrained Vision-Language-Action models.
Unknown
This paper investigates reinforcement learning for interactive tool-using agents, highlighting user model fine-tuning as a critical component for training such agents.
Guiyao Tie, Zeli Zhao, D. Song, et al.
First comprehensive survey of post-training LLMs, covering fine-tuning, alignment, reasoning, efficiency, and integration, with a taxonomy and agenda for Large Reasoning Models.
Zeyang Sha, Shiwen Cui, Weiqiang Wang
SEM is a post-training RL framework that explicitly trains LLMs to optimize search usage, reducing redundant search while maintaining or improving answer accuracy.
Unknown
Opsdl introduces an on-policy self-distillation method to extend the effective context length of LLMs through post-training.
Yaoming Li, Guangxiang Zhao, Qilong Shi, et al.
First primer synthesizing over 150 studies on post-training reasoning data, organizing the field around four key questions.
Tanishq Kumar, Zachary Ankner, B. Spector, et al.
Proposes precision-aware scaling laws for language models, showing low precision reduces effective parameter count and post-training quantization can harm models trained on more data.